Shaowei Xia

Learn More
Mining association rules from large databases is an important problem in data mining. There is a need to develop parallel algorithm for this problem because it is a very costly computation process. However, all proposed parallel algorithms for mining association rules follow the conventional level-wise approach. On a shared-memory multi-processors, they(More)
Abstiact W'e present a new probtilfisdc cl=-fier, ded SOM-PNN class-tier, for volume data class-fi~on md ~om The new classifier produces probabilistic cl=-fication with Bayesian cofidenu measure which is hi@y desirable in volume rendering Based on the SOM map traind with a large training data set our SOM-P~ classifier performs the probabiic classification(More)
This paper presents two compensation methods for multilayer perceptrons (MLPs) which are very difficult to train by traditional Back Propagation (BP) methods. For MLPs trapped in local minima, compensating methods can correct the wrong outputs one by one using constructing techniques until all outputs are right, so that the MLPs can skip from the local(More)
This paper presents a new scrambling algorithm forimage encryption. It combines the Logistic chaotic sequence andthe common Rubik's Cube. The algorithm partition an originalimage into several blocks and generates some cubes. Rotate thesecubes similar to Rubik's Cube by using the methods which aregenerated by Logistic system. After that, we use those cubesto(More)
This paper introduces a novel image scrambling algorithm based on Sudoku puzzle. According to the special property that every number from 1 to N appears only once in each row or column in an N*N Sudoku puzzle, a 1-1 relationship can be setup between two Sudoku puzzles and these two Sudoku puzzles will be used to map the original images to a scrambled one.(More)
In this paper, we produce a new medical image classification scheme using self-organizing map (SOM) combining with multiscale technique. It addresses the problem of the handling of edge pixels in the traditional multiscale SOM classifiers. First, to solve the difficulty with manually selection of edge pixels, a multiscale edge detection algorithm based on(More)
This paper improves an image scrambling algorithm based on Rubik's cube rotation and logistic sequence. The improved algorithm resizes original image and partitions a resized image into six blocks to generate cubes. Rotate these cubes in 25 steps by using 30 different rotating methods which are controlled by a chaotic system. After rotating, remap those(More)